PHYSICAL PROPERTIES OF WATER, ETC. 13 



Thus water appears to have its minimum compressibility (for low pressures) about 

 63° C. 



My own earlier determinations 1 will be given more fully below (Section VI.). I 

 may here quote one or two, premising that they were given with a caution (not 

 required, as it happens), that the pressure unit of my external gauge was somewhat 

 uncertain. They are true, not average, compressibilities. See Appendix B. 



At 12°-0 C. 



Fresh water 0-00720 (1 - 0-03-ip) 



Sea water 0-00666 (1 - 0-034^) 



At 15°-5 C. 



Fresh water 0-00698 (1 - 0-05^/) 



Sea water 0-00645 (1 - 0-05i') 



Ratio 

 1 : 0-925 



Ratio 

 1 : 0-924 



In all of these the unit of pressure is one ton-weight per square inch (152'3 atm.). The 

 diminution of compressibility with increased pressure was evident from the commence- 

 ment of the investigations. I assumed, throughout, for the compressibility of glass 



0-000386 per ton, 

 which, as will be seen below, is a little too small. 



By direct comparison with Amagat's manometer, I have found that the pressure 

 unit of my external gauge is too small, but only by about - 5 per cent. This very slight 

 underestimate of course does not account for the smallness of the pressure term of the 

 first expression above. As will be seen later, the true cause is probably to be traced to 

 the smallness of the piezometers which I used in my first investigations, and to the fact 

 that their stems were cut off " square " and dipped into mercury. Allowing for this, it 

 will be seen that the above estimates of compressibility agree very fairly, in other 

 respects, with those which I have since obtained. The sea-water employed in the 

 comparison with fresh water was collected about a mile and a half off the coast at 

 Portobello, and was therefore somewhat less dense (and more compressible) than the 

 average of ocean-water. In my later experiments, to be detailed below, the sea-water 

 operated on was taken at a point outside the Firth of Forth, considerably beyond the 

 Isle of May. 



As stated in my Beport on the Pressure Errors, &c, the unit of my external 

 gauge was determined by the help of Amagat's data for the compression of air. As the 

 piezometer containing the air had to be enclosed in the large gun, the record was 

 obtained by silvering the interior of the narrow tube into which the air was finally 

 compressed : — and the heating of the air by compression, as well as the uncertainty of 



1 Proc. Roy. Soc. Edin. 1883 and 1884. 



